Recommend two papers

[realwangjiahao]

Thank you for summarizing these articles, it is very helpful for my research work.
I recently read two papers on using large language models for time series classification. These two articles inspired me a lot, I recommend them to you here.

The first one is "InstructTime: Advancing Time Series Classification with Multimodal Language Modeling", and this article proposes a new method called InstructTime, which uses gpt-2 to encode time series and achieves better classification results. You can find the article at https://arxiv.org/abs/2403.12371. And code can be used at https://github.com/Mingyue-Cheng/InstructTime.

The second paper is “CrossTimeNet: Learning Transferable Time Series Classifier with Cross-Domain Pre-training from Language Model.”
This article uses BERT and GPT-2 to encode time series and uses self-supervision methods to improve classification performance. You can find the article at https://arxiv.org/pdf/2403.12372. And code can be found at https://github.com/Mingyue-Cheng/CrossTimeNet.

These two articles can be divided into the "PFMs for Time Series" part。

Could you please add the two papers to your summary list? Thanks a lot.